Infrastructure for Prevention 1439
ystem discussed in Chapter 2.) Main (M) emphasis codes are used only or parent projects of multiproject awards and are equivalent to the >rimary (P) emphasis level. Therefore, in the committee's review, main nd primary levels were combined.
All items that appear in a CRISP printout are peer reviewed when hey are proposals. However, the type of peer review required varies .ccording to whether the monies are intended for a grant or a contract. ^. contract proposal is not reviewed by the same standards or mecha-dsms as a grant proposal. The contract proposals are reviewed through
process called a secondary review, in which the review committee is omposed of a certain percentage of NIH personnel and a certain •ercentage of outside experts.
The difficulty in assessing the funding commitments by federal gencies to prevention research is dramatized by the CRISP recording ystem at NIH. The CRISP system has a number of serious problems /hen used to evaluate research programs. First, when funded projects ome into the system, they are classified by a staff person. Classifica-ions are not consistent, in part because a written definition of preven-ion has not been available for guidance. Second, the data are not omplete. For example, the CRISP lists of research projects did not natch the portfolios of funded grants supplied by the relevant program •fficers at NIMH and NIDA. Only 24 of 80 grants funded from 1988 to 992 by the NIMH Prevention Research Branch (PRB) were even listed »y CRISP as having a main/primary area of emphasis on prevention, 'hird, the same project can be classified in more than one prevention rea, for example, alcoholism prevention and mental disorder preven-ion. Fourth, there is no limitation on how many times a particular •reject can be classified by emphasis code. For example, in fiscal year FY) 1991, 41 research projects (which were funded by NIMH, NIDA, •JIAAA, the National Institute of Child Health and Human Development (NICHD), and the National Center for Research Resources MCRR)) were identified as having a main/primary emphasis in mental Lisorder prevention. The projects totaled $13,675,607. The numbers of »ther primary, secondary, and tertiary areas of emphasis that each •articular grant was classified as having were counted. These included reas as diverse as "psychometrics" to "social control" to "immunopa-hology." The 41 projects had 264 primary areas of emphasis in addition o mental disorder prevention, 312 secondary areas, and 114 tertiary reas.
The committee concluded that the CRISP system can miss directly elevant projects, on the one hand, and also can produce extensive nultiple counting of projects on the other, which can result in mislead-d computerized lists are available.s in the bibliography. The entries totaled 168, but researcher accounted for 28 of these. Many NIMH PRB researc including those whose projects are completed, have never publi their findings in peer-reviewed academic journals; others have wi book chapters that contain some data.